Speech Emotion Analysis and Recognition Based on the PCA Feature Extraction Model

Abstract

With the development of human-computer interaction technology, research on man-machine interface has stepped from the mechanization into the era of multimedia user interface. As one of the key problems of intelligent human-computer interaction technology, speech emotion analysis and recognition has attracted more and more attention. In view of that, this paper presents a speech emotion recognition model based on PCA feature extraction model. It firstly introduces the speech emotion recognition system and the pretreatment process of the speech signal, including anti-aliasing filtering, pre-emphasis, windowing and framing and endpoint detection etc, and then proposes an improved recognition model based on PCA feature extraction model, substituting the original variables with a group of new independent variables and choosing several main variables that can represent the original variable from the new variables, so as to improve the accuracy of problem solving and simplify the computational complexity. The simulation results show that, compared with the traditional algorithm model, the proposed speech emotion recognition model based on PCA feature extraction model reduces the classification noise of uncelebrated corpus and increases the average recognition rate of the system.